Between December 1st 1997 and November 30th 1998 there were no new subjects enrolled into this study. Four follow up studies were performed. The last follow up study was completed by March 1998, and data analysis is now (as of February 1999) complete. Based on this pilot study of the growth and nutrition of children with intractable epilepsy on the ketogenic diet we conclude the following. Preliminary Studies Aim 1: To determine the efficacy of the standard 4:1 ketogenic diet (KD) As part of a pilot study of the growth and nutritional status of children on the currently recommended 4:1 KD preceded by a fast, efficacy was measured. Twenty-five subjects with intractable epilepsy were enrolled over a two-year time interval and followed prospectively for 6 months. Follow-up evaluations were scheduled at 1, 3 and 6 months. We defined efficacy as >50%reduction of baseline seizure frequency. Change in seizure frequency was tabulated from the parent's daily seizure calendars and the average seizure frequency (number of seizures per week) occurring during 28 days before the 3-and 6-month follow-up visits were compared with baseline. The percentage decrease in seizures from baseline was used to assess the efficacy of the diet. Adverse events were also noted. Aim 2: To determine if the restrictive nature of the KD places the children at risk for nutritional deficiencies that will impair growth rate and accretion of fat-free mass. The second project was a longitudinal study of growth and nutritional status of children with intractable epilepsy treated with the KD. The same 25 subjects were followed at baseline, 1-month, 3-months and 6-months on the KD treatment. Weights and heights were obtained and measurements were converted to Z scores. Body composition was obtained using the dual energy x-ray absorptiometry (DXA) and expressed as fat-free mass (FFM) (lean mass + bone mass) and % body fat. Aim 3: To determine if children on the ketogenic diet are at risk for osteopenia. It is well known that anticonvulsants may predispose children to osteopenia through alterations in vitamin D and calcium metabolism. Whether the KD worsens this effect has not been well studied. We evaluated bone health in 14 of the 25 children by measuring Lumbar AP spine bone density (BMD) and bone mineral content (BMC) before the KD and at 3 months into the maintenance diet. Aim 4: To determine if change in substrate utilization as measured by the respiratory quotient (RQ) can predict who will respond to treatment with the ketogenic diet. The mechanism of the KD remains unknown. The currently recommended KD protocol at CHOP requires a 3 month commitment (because of some late responders) before efficacy can be determined. A predictor of response would simplify the management of these patients. We measured Resting Energy Expenditure (REE) and RQ in the previously described 25 subject to see if changes in substrate utilization as measure by the RQ could predict response to KD.